bluetooth BMS?

ElectricGod said:
IE: C- mosfets have their diodes reverse biased when charging from C-. From P-, current flow into the BMS if the mosfets are off is always possible since the back diodes are forward biased. However current flow out of the BMS if the mosfets are off is stopped since the back diodes are reverse biased. Effectively what this means is from P- you can't stop charging current only discharging current flow.
Thats exactly what i was trying to say with "current can flow always in one direction". It seems that it wasn't well expressed.
The C- FET's cannot switch off a load (or doing low voltage cut off), and the P- FET's cannot disconnect a charger (or doing high voltage cut off).
If the current is flowing through the body diode there will be a larger voltage drop (around 1V instead of the usual few mV), but current continuous to flow if it gets turned off (or with 0V on the Gate).
So then, lets talk about drastically BAD cell voltages for a minute. I'm sure NOT perfect and sometimes I do stupid stuff and kick myself really hard for it. This was at 12S and 48 volts. I was a noob to EV's. I had built 4 12S LIPO packs and was using a crappy charger that topped out at 52 volts. That's 2 volts over 4.2 per cell or 50.4 volts. This was not ideal for my LIPO packs and I had to watch that the voltage at the EV never went above 50.4 volts...or at least not by much. Charging over night was out of the question as all of the current from the charger would get spent in lots of balance circuits among 4 BMS in parallel. Fortunately the charger was capable of 2 amps so this wasn't a huge concern if I was watching the charge process periodically. I had purchased 3 BMS and they were in active use, but my 4th pack was waiting for it's BMS to arrive. It was common that I would plug in the charger into the EV and a couple of cell monitors into the unprotected pack. When the loud beeping started, it was time to pull the unprotected pack and finish charging via my RC charger. For whatever dumb reason, I flat forgot to do this. As you can imagine, that 12S pack was sitting at 52 volts or 4.33 volts per cell when I realized my mistake. I was thinking to myself why wasn't I hearing the cell monitors? I immediately realized it was becasue I had never plugged them in. I ran to the scooter and sure enough, every cell in that pack was at 4.3 volts and the pack was quite hot. I pulled it from the scooter, left it in the middle of the floor in my garage and let it cool down. All the cells now showed 4.4 volts!!! I then attached a load and pulled them back down to 4.1 volts per cell. That pack did suffer some loss of capacity, but no cell exploded or even puffed up. 2 weeks later it had it's own BMS. The other 3 12S packs that had BMS's were just fine. The cells despite the charger producing too much voltage were all sitting at 4.2 volts and soon were back down to 4.1 volts thanks to the BMS's.

Glad to hear that there was no fire^^
I am absolutely with you that it needs very bad cell voltage differences to overcharge one group significant so that there will be a fire.
But think what happens if somebody has one 10s charger and one 13s charger with same charging plug and they get confused.
It would charge to >5,4V per cell and booooom bang fire, because the BMS cannot turn it off.
 
madin88 said:
Thats exactly what i was trying to say with "current can flow always in one direction". It seems that it wasn't well expressed.
The C- FET's cannot switch off a load (or doing low voltage cut off), and the P- FET's cannot disconnect a charger (or doing high voltage cut off).

IMHO, this is a worth while trade off. Set your regen voltage correctly and you can't run voltage into your pack higher than it is capable of handling. Regen set to max pack voltage can't push more than max pack voltage into the pack which is completely harmless. Not that this is likely anyway even with regen set too high. It would take going down a mile long hill on a fully charged pack. 50 feet down a hill with regen set too high on a charged pack won't make a difference. The time frame is way too short to matter. Set your charger and regen to 4.1 volts x cell count and then you can't over charge. This eliminates the need for over voltage (C-) mosfets. THat's their only real purpose for existing.

madin88 said:
Glad to hear that there was no fire^^
I am absolutely with you that it needs very bad cell voltage differences to overcharge one group significant so that there will be a fire.
But think what happens if somebody has one 10s charger and one 13s charger with same charging plug and they get confused.
It would charge to >5,4V per cell and booooom bang fire, because the BMS cannot turn it off.

I tried that once...I was really surprised how much voltage it took for the lipo to over heat and catch fire. You hear all the LIPO/LION FUD about them catching fire and I have to say it's 90% BS. They don't spontaneously explode. They have to get pretty hot first. I wish when I tried blowing up LIPOs that I had a laser temp meter at the time, but didn't. Anyway...4.3 volts got the batteries pretty warm, but not enough to cause any swelling and I could hold the cells in my hand. Same for 2 6S 12,000mah LIPO's I recovered. They were at 149F I think...that's like a furnace to LIPO! Nothing happened despite being ran down to nearly dead flat. One cell puffed up. The rest of the cells I recovered and I'm still using those packs 2 years later. One of the 6S packs is a 5S pack, but so what...the rest of the cells suffered very little capacity loss. Anyway, back on my "burner LIPOS" I think I got to 4.6 volts before one finally caught fire. By then, it was way to hot to hold in my hand. AND these were really cheap LIPO's. I also drove a nail rough a fully charge burner LIPO. Some small amounts of smoke curled around the nail hole and that was all. No fire, no explosion...I was really disappointed!

It happens that people do stupid things. Count me in! I've done my fair share!!! However, that's IMHO an operator error, NOT a product error. Anything fool proof can be overcome by a sufficiently stupid fool.

If it bothers you, charge at C- and run the EV at P-. I personally would not do more than that. You get the over voltage protection when charging and everything else gets the LVC and over current protection from the P- mosfets without the inline series losses of the C- mosfets.

IMHO...this is bad and I did it. discharging out of C-...UGG!!! In my copious free time, I'll take these BMS back off the packs and fix this oversight.

16S%20Yimia%20LION%20BMS%20complete_zpszonxjund.jpg
 
What happens when discharge current is exceeded?
I don't know as I have been afraid to trust the p- Mosfets. :oops:
Will Mosfets turn off and bms will need reset?
Will it damage Mosfets?
Let's assume that the bms current is set properly.
 
Inwo said:
What happens when discharge current is exceeded?
I don't know as I have been afraid to trust the p- Mosfets. :oops:
Will Mosfets turn off and bms will need reset?
Will it damage Mosfets?
Let's assume that the bms current is set properly.

The mosfets can handle more current than the BMS is rated for...at least in the short term. Since they have no contacts to make or break, there is no arcing. As the mosfets start turning off their internal resistance rises which limits current flow. Depending on the specific mosfet, fully on to fully off can be milliseconds to nano seconds. Less time is generally better in BMS and motor controllers. In general, turning off the mosfets, just cuts current and that's that. You do this thousands of times every minute in a motor controller under load. The exact same process happens in a BMS, but only in over load situations and on to off or off to on happens only once.

Whether you like it or not you are always trusting the P- mosfets. They are there for over current and LVC shut off. When charging, you are using the P- mosfets. In order for a BMS to work, you are 100% of the time using the P- mosfets.
 
Inwo said:
What happens when discharge current is exceeded?
I don't know as I have been afraid to trust the p- Mosfets. :oops:
Will Mosfets turn off and bms will need reset?
Will it damage Mosfets?
Let's assume that the bms current is set properly.

Most BMS will have overcurrent protetcion (measured over a shunt or ON resistance of the FET's), but this OCP is many times higher as the rated current.
A properly build BMS should have temperature monitoring with one sensor on the board, and a second one on a wire which can be installed between the cells.
 
where is the best place to put a switch to disable this BMS from long term storage, is the B- negative the best place to stop parasitic drain ? it does not draw anything from the balance leads does it ?

Also if i have two 36v batteries with 10S BMS, will their be an issue if i put the batteries in series and charge with a 72v bulk charger ? as each battery only sees 36v ?
 
Alex07 said:
where is the best place to put a switch to disable this BMS from long term storage, is the B- negative the best place to stop parasitic drain ?
K1 connector if you have it. BMS is powered by balance wires. So to completely disable it you will have to disconnect it completely.
 
Alex07 said:
where is the best place to put a switch to disable this BMS from long term storage, is the B- negative the best place to stop parasitic drain ? it does not draw anything from the balance leads does it ?

Also if i have two 36v batteries with 10S BMS, will their be an issue if i put the batteries in series and charge with a 72v bulk charger ? as each battery only sees 36v ?

Be sure to run your batteries down to about 3.7-3.8 volts per cell if you intend to store them long term. and of course completely disconnect the BMS from the batteries.
 
circuit said:
Already tried this?
https://www.energusps.com/shop/category/battery-management-4

Three options available: 30, 150 and 750 amps. Up to 16s.

yes, they are my default choice for any battery that uses 16S or less. quality stuff as long as you use the proper isolated communication cable (you will blow the com lines if you charge it and hook it up to a computer/laptop that is plugged in)

i do recommend getting the 750A meter if you go beyond 40~50A continous load. there are 2 very nice mosfets but you will start burning up power on those fets and the board needs to be cooled. i never had issues with them but if you are going to build a serious high current battery (under 16S) then this is the one to get.
otherwise you might still use this one and daisy chain them on a longer pack. but the cost gets nasty when you need to order 7~15 euros worth of wires SEPARATLY in order to get the bloody thing working. thats pushing my buttons on a 150 euro board with 30 euro worth of compontents on it.

right now i have a need for more then 16S so i am trying my luck with the oriental style bms found here.

i have 3 right now that all are awating a new home for batteries i am making for customers. so i hope these 120A versions i orderd can handle 30~50A for the next few years. leat least until the warranty is far gone, i need to make a reputation.
 
flippy, you probably did not notice it's my company :)
I am glad you like it.

By the way, there is a hack to get up to 150A constant without overheating. You can simply bypass the mosfets and connect your motor controller right after onboard current sensor, and configure cutoff device from mosfet to high side switch (one of programmable outputs). You can then wire your controller's "logic" supply to that output. In event of undervoltage, the BMS will cut off that logic part of your controller and it will be the same as complete cutoff.
 
circuit said:
flippy, you probably did not notice it's my company :)
I am glad you like it.

By the way, there is a hack to get up to 150A constant without overheating. You can simply bypass the mosfets and connect your motor controller right after onboard current sensor, and configure cutoff device from mosfet to high side switch (one of programmable outputs). You can then wire your controller's "logic" supply to that output. In event of undervoltage, the BMS will cut off that logic part of your controller and it will be the same as complete cutoff.

no, i did not notice that.

but i do stand by my statement.

it is a lovely product but for the price not supplying it with basic wires is rubbing me the wrong way.

sidenote: please make the wires 20cm longer or so. its not hurting the bottom line and often can prevent having to exend 16 wires. if one has to build these things into batteries several times a month its a tad annoying and takes a lot of time. for the price one can expect a fair length, or sell a kit with longer wires.
 
Thank you for valueable feedback. This is not a first time I hear about wire length, so I'll make sure we will be able to offer longer balancing wire kits soon (couple months?).

By the way, there were several firmware updates, check release notes. Contact support for update file.
 
circuit said:
Thank you for valueable feedback. This is not a first time I hear about wire length, so I'll make sure we will be able to offer longer balancing wire kits soon (couple months?).

By the way, there were several firmware updates, check release notes. Contact support for update file.

It sure would be nice if you made BMS larger than 16S...such as at least 20S and realistically 32S. I'm sure your products are great, but 16S is just way too low of a cell count for lots of things.
 
ElectricGod said:
circuit said:
Thank you for valueable feedback. This is not a first time I hear about wire length, so I'll make sure we will be able to offer longer balancing wire kits soon (couple months?).

By the way, there were several firmware updates, check release notes. Contact support for update file.

It sure would be nice if you made BMS larger than 16S...such as at least 20S and realistically 32S. I'm sure your products are great, but 16S is just way too low of a cell count for lots of things.

Or find a way to get two or more to work together in series. Perhaps an rs485 or rs422 network. If you could accomplish that I believe 16s as a building block would be a good start.

Granted I don't know an awful lot about this stuff but I am learning.
 
in theory just having a balance board with a data connection would be enough. it would not require the full bms. a fairly simple board (no mosfets or other crap) that just measures voltages and relays that to the main board would be enough and cheaper.

buying 300+ euro for a dual bms option is hard to sell (for me as a company).
as i am one basically the only one in my country that can build custom batteries for the govecs brand and can modify the sevcon controller means i get more and more people with a need for 20S (and higher) battery and changing the voltage (from lead to lithium for example) requires changing settings in the sevcon and i am one of the few in europe that can do this.
it sucks i dont have a cost effective european made bms to offer. i dont like dealing with china if i can source locally.

right now the 100 euro cost of a 32S smart bms as discussed in this topic is much better to swallow for customers. especially as that has a lcd and blutooth as standard. guess what people choose when i give them the option for either strapping 2 boards together wich adds up to over 400 euro's for a usb/bluetooth versus 100 euro. ofcourse there is a difference in quality but even taking a 30 minute solder job to replace the mosfets with decent stuff on the chinese board still gives me a 250 euro advantage and the customer gets a nice lcd screen to play with and a free app on their phone. i cant fault customers for never opting for dual a tinybms solution. i only use the tinybms now for standard 48V conversions of scooters. anything above that requires a more capable BMS.

having a tinbmsXL would solve many issues, still one board but with 32S balancing and the same mosfet sulution. only needs a bit taller board for balancing resistors and connections and prehaps a second atmel for measuring voltages. should be doable and combined with a 750A option it becomes a really powerful option for large setups and proper EV vechicles.
 
flippy said:
in theory just having a balance board with a data connection would be enough. it would not require the full bms. a fairly simple board (no mosfets or other crap) that just measures voltages and relays that to the main board would be enough and cheaper.

buying 300+ euro for a dual bms option is hard to sell (for me as a company).
as i am one basically the only one in my country that can build custom batteries for the govecs brand and can modify the sevcon controller means i get more and more people with a need for 20S (and higher) battery and changing the voltage (from lead to lithium for example) requires changing settings in the sevcon and i am one of the few in europe that can do this.
it sucks i dont have a cost effective european made bms to offer. i dont like dealing with china if i can source locally.

right now the 100 euro cost of a 32S smart bms as discussed in this topic is much better to swallow for customers. especially as that has a lcd and blutooth as standard. guess what people choose when i give them the option for either strapping 2 boards together wich adds up to over 400 euro's for a usb/bluetooth versus 100 euro. ofcourse there is a difference in quality but even taking a 30 minute solder job to replace the mosfets with decent stuff on the chinese board still gives me a 250 euro advantage and the customer gets a nice lcd screen to play with and a free app on their phone. i cant fault customers for never opting for dual a tinybms solution. i only use the tinybms now for standard 48V conversions of scooters. anything above that requires a more capable BMS.

having a tinbmsXL would solve many issues, still one board but with 32S balancing and the same mosfet sulution. only needs a bit taller board for balancing resistors and connections and prehaps a second atmel for measuring voltages. should be doable and combined with a 750A option it becomes a really powerful option for large setups and proper EV vechicles.

That's basically where I was going with my last comment, just didn't quite make it being half asleep.

The TinyBMS, on paper at least as I have not actually used one yet, has the potential to be a great, and reasonably priced, product in a much larger market if they go that route. A Master unit that can have N slaves for additional cell management. That would be very close to how the BECM system in the Chevy Volt/Opel Ampera works. It has a master, the BECM1, that interfaces with cell managment boards over CANBus. It handles the remaining functionality like pack monitoring, HV disconnect, fault reporting, charge interface, etc.
 
Monstarr said:
flippy said:
as i am one basically the only one in my country that can build custom batteries for the govecs brand and can modify the sevcon controller

:roll:
i said basically. :wink:
 
I fully understand the need of BMS for >16s packs, we get this a lot. There are several topologies out there:
* Centralized (TinyBMS)
* Distributed (for example, EMUS BMS)
* Mixed / Star topology. Which is basically central control unit + several centralized boards.

Each of these have specific markets and applications. We had a choice: try to do it all, or do one, and do it well. And we chose to do well.
Furthermore, >16s is mainly used in DIY builds, and also some weirdly designed solutions, like Zero and several others. Industry and the "big fishes" are sticking with standard voltages, like 24V, 36V and 48V.
We could do a "special" solution for DIYers, but let's be honest here: this market is very tough due to high demand for after-sale tech support. Basically you sell product for 200 € and then have to exchange 20 emails with highly technical Q&As, which is usually done by skilled engineer, as typical sales guys can't dive that deep.

So for now I can say this: we are not likely to do >16s board soon, however we are working on star topology solution for systems up to 2000V. But that will be likely out of DIY price range.
 
I fully understand circuit explanation. The biggest problem with battery voltage is in the legislative. In EU there is limit of 60 VDC as extra-low voltage which is considered to be safe aginst electric shock under all conditions. If you want to legaly sell the batteries with higher voltage than 60 VDC, you must meet stricter regulations (e.g. isolated HV wiring with ground fault detection, orange marking of HV conductors and connector in automotive) and usually perform several expensive tests. So if your application requires higher operational voltage than 60 VDC, there is no difference between for example 120 VDC and 400 VDC in terms of legislation. The 120 VDC system components are only significantly cheaper than 400 VDC, especially in low-rate production and for DIYers.

From my point of view this is the reason why there are exist many "up to 16s" professional BMS on the market, but few in the range of 24-28s
 
Well said, Pajda.

In addinion, I want to stress that you don't actually need to go above 48V in ebike applications, at least below 10-20kW or so. I myself had several DIY ebikes that could classified as "fast" (up to 80km/h) and all of them run on 48V. 16s LiFePO4 in 2009, and 12s Li-ion today.

Why are people going this high with voltage? Motor too slow? It's a small problem, compared to all the problems you face with higher voltage: lack of quality BMS, problems finding quality charger, very few non-china controllers available.

Just select a proper motor winding and all your problems will go away. For example, there are very compact, cheap and good quality chargers from meanwell and eltek. You will find nothing like that above 60V.
 
ES is what... 80-90% DIYers? Industry essentially sells safe and slow and weak EV's for ridiculously high prices. $1400 for a kick scooter that barely goes 20mph...LAME! There's reasons why there's DIYers. We want speed and good acceleration and good range and that does NOT exist on the commercial market without paying a fortune for it. I can roll my own, have loads of fun building it and still be 40% or less the cost of something commercially made that performs similarly. The math just doesn't work out to buy a ready built EV IMHO.

all of that to say this...
ES is all about do it yourself builds. So when we want something, there's reasons for it. 16s has it's place. My favorite kick scooter runs at 16S and 60 amps and does 45mph. I have less than $500 in it. That IS my "low voltage build...66 volts"! Every other build I've ever done starts at 20S and I want more voltage, not less in the future. 20S is just barely ticking the corner of the box for me.

I'm not faulting the tinyBMS for any of this and frankly a good BMS is worth the money. BUT, when considering that China produces "good enough" battery tech, most people are going to make the choice to go with "good enough" and even more so if it has the cell counts we want.

Hey Circuit...
I have an idea for you. Your BMS is an all in one device. How about split out the power mosfets to a separate board and then make the balancing and monitoring boards stack-able. This allows your product to reach a far broader market (people like me!!!) where 16S is flat out a show stopper. The larger my pack is, the more it costs and the more I'm willing to pay for battery protection. At 16S, IMHO, I'm going to go with whatever is cheap for no other reason than the battery pack cost is not that great. IE: It's NOT compelling to spend a lot of money on a BMS that costs almost as much as my batteries do. With a stack-able board option and separate mosfet board, you can build 16S BMS all you want and then expand it to a 32S option that works at 20-32S. Add a third balance board and now you've reached 95% of the DIY market and probably a lot of the high end commercial builders too! The mosfet board can then be built in several options...such as 50, 100, 200, 300 amps. And of course build them to voltage requirements. The Chinese do this all day long. Why can't Western world BMS makers figure this out and then do it for a reasonable price?
 
As mentioned before, you don't actually need mosfets in the BMS, you can use logic power cutoff to achieve the same result without bulky mosfets and heat. If done this way, there is no current limit.


Please help me to understand this. What is your reasoning for going above 60V? Speed? You can have that with 48V, easily.
 
there is very good reasons to go in higher voltages. first: much less amps. that means less peukert effect rubbish, thinner cables. less voltage loss and so on.

personally the main problem is that most scooters and stuff is made to run 40km/h and if the battery is full prehaps 45 for a few minutes slowly dipping into the high 30's when getting empty. that does not ride well or sell very well. it's also dangrous in the city being slower then cars. people will make dangerous overtakings just to get past you.
many people want more speed and keep that speed even if the pack discharges. so keeping 50km/h at all times requires basically a 72V battery or 20S instead of 16S.
replacing the original motor for a new one from QS is a expensive ordeal and usually means a lot of crap getting the rear brake to fit properly. adding the cost of a new motor, controller and battery makes it very cost prohibitive and most people will simply walk away from this. being able to reuse the existing motor makes a huge difference. but you do need the higher voltage to get the speed people expect across the entire voltage range.
 
Back
Top